9:05 p.m.
Hi,
another minor question before release. Consider this example:
<hansl>
strings S = defarray("a")
print S[i] # no output, no reaction
</hansl>
I _don't_ expect this to work, that's not the point. But shouldn't there
be some kind of parsing error?
thanks
sven
12:29 a.m.
On Sun, 10 Jan 2021, Sven Schreiber wrote:
Hi,
another minor question before release. Consider this example:
<hansl>
strings S = defarray("a")
print S[i] # no output, no reaction
</hansl>
I _don't_ expect this to work, that's not the point. But shouldn't there
be some kind of parsing error?
Yes. Fixed.
Allin
4:24 p.m.
Am 11.01.21 um 01:29 schrieb Allin Cottrell:
On Sun, 10 Jan 2021, Sven Schreiber wrote:
> Hi,
> another minor question before release. Consider this example:
>
> <hansl>
> strings S = defarray("a")
> print S[i] # no output, no reaction
> </hansl>
>
> I _don't_ expect this to work, that's not the point. But shouldn't there
> be some kind of parsing error?
Yes. Fixed.
Thanks for the quick fix, Allin. However, I expected to be able to catch
the error by
<hansl>
strings S = defarray("a")
catch print S[i]
eval $error
</hansl>
but I get the error message:
"> S[i]
The symbol 'i' is undefined"
Artur
7:21 p.m.
On Mon, 11 Jan 2021, Artur Tarassow wrote:
Am 11.01.21 um 01:29 schrieb Allin Cottrell:
> On Sun, 10 Jan 2021, Sven Schreiber wrote:
>
>> Hi,
>> another minor question before release. Consider this example:
>>
>> <hansl>
>> strings S = defarray("a")
>> print S[i] # no output, no reaction
>> </hansl>
>>
>> I _don't_ expect this to work, that's not the point. But shouldn't
there
>> be some kind of parsing error?
>
> Yes. Fixed.
Thanks for the quick fix, Allin. However, I expected to be able to catch the
error by
<hansl>
strings S = defarray("a")
catch print S[i]
eval $error
</hansl>
but I get the error message:
"
You're getting a warning, no?
<hansl>
strings S = defarray("a")
catch print S[i]
eval $error
print "still going!"
</hansl>
<output>
? strings S = defarray("a")
? catch print S[i]
S[i]
The symbol 'i' is undefined" S[i]
The symbol 'i' is undefined
? eval $error
2
still going!
</output>
Allin
7:28 p.m.
Am 11.01.21 um 20:21 schrieb Allin Cottrell:
On Mon, 11 Jan 2021, Artur Tarassow wrote:
> Am 11.01.21 um 01:29 schrieb Allin Cottrell:
>> On Sun, 10 Jan 2021, Sven Schreiber wrote:
>>
>>> Hi,
>>> another minor question before release. Consider this example:
>>>
>>> <hansl>
>>> strings S = defarray("a")
>>> print S[i] # no output, no reaction
>>> </hansl>
>>>
>>> I _don't_ expect this to work, that's not the point. But
shouldn't
>>> there
>>> be some kind of parsing error?
>>
>> Yes. Fixed.
>
> Thanks for the quick fix, Allin. However, I expected to be able to
> catch the error by
>
> <hansl>
> strings S = defarray("a")
> catch print S[i]
> eval $error
> </hansl>
>
> but I get the error message:
> "> S[i]
> The symbol 'i' is undefined"
You're getting a warning, no?
<hansl>
strings S = defarray("a")
catch print S[i]
eval $error
print "still going!"
</hansl>
<output>
? strings S = defarray("a")
? catch print S[i]
> S[i]
The symbol 'i' is undefined
? eval $error
2
still going!
</output>
Interesting, indeed it works but only outside a function. I ran this
within a function where it stops hard:
<hansl>
set verbose off
clear
set assert stop
function void test_print_strarray_unvalid_idx (void)
print "Start testing print command."
# Given
strings S = defarray("a")
scalar err_expected = TRUE
# When
catch print S[i]
scalar err_actual = $error
print err_actual
print "still going!"
# Then
assert(err_actual == err_expected)
end function
test_print_strarray_unvalid_idx()
</hansl>
Artur
8 p.m.
On Mon, 11 Jan 2021, Artur Tarassow wrote:
Am 11.01.21 um 20:21 schrieb Allin Cottrell:
> On Mon, 11 Jan 2021, Artur Tarassow wrote:
>
>> Am 11.01.21 um 01:29 schrieb Allin Cottrell:
>>> On Sun, 10 Jan 2021, Sven Schreiber wrote:
>>>
>>>> Hi,
>>>> another minor question before release. Consider this example:
>>>>
>>>> <hansl>
>>>> strings S = defarray("a")
>>>> print S[i] # no output, no reaction
>>>> </hansl>
>>>>
>>>> I _don't_ expect this to work, that's not the point. But
shouldn't there
>>>> be some kind of parsing error?
>>>
>>> Yes. Fixed.
>>
>> Thanks for the quick fix, Allin. However, I expected to be able to catch
>> the error by
>>
>> <hansl>
>> strings S = defarray("a")
>> catch print S[i]
>> eval $error
>> </hansl>
>>
>> but I get the error message:
>> "> S[i]
>> The symbol 'i' is undefined"
>
> You're getting a warning, no?
>
> <hansl>
> strings S = defarray("a")
> catch print S[i]
> eval $error
> print "still going!"
> </hansl>
>
> <output>
> ? strings S = defarray("a")
> ? catch print S[i]
>> S[i]
> The symbol 'i' is undefined
> ? eval $error
> 2
> still going!
> </output>
Interesting, indeed it works but only outside a function. I ran
this within a function where it stops hard [...]
I'm going to leave this one till after release. "print" is kind of
an oddity; you should find that catch works OK with "eval" instead.
Later we can consider merging eval and print, so that (a) print
works for sub-objects, as eval does, but (b) it does the current
"full dress" print job if the object turns out to be a top-level
variable.
Allin
10:11 p.m.
Am 11.01.2021 um 21:00 schrieb Allin Cottrell:
On Mon, 11 Jan 2021, Artur Tarassow wrote:
> Interesting, indeed it works but only outside a function. I ran
this
> within a function where it stops hard [...]
I'm going to leave this one till after release. "print" is kind of an
oddity; you should find that catch works OK with "eval" instead.
Later we can consider merging eval and print, so that (a) print works
for sub-objects, as eval does, but (b) it does the current "full dress"
print job if the object turns out to be a top-level variable.
and ..... this is now a ticket:
https://sourceforge.net/p/gretl/feature-requests/136/
thanks
sven
1442
days inactive
1443
days old
6 comments
3 participants
participants (3)
-
Allin Cottrell -
Artur Tarassow -
Sven Schreiber